The constant-Q theory for anelastic attenuation is reviewed, particularly the aspects related with the determination of the phase of the attenuated signal and the potential error introduced by the discrete implementation of the Hilbert transform. Synthetic data are used to test the phase correction in the Gabor deconvolution and to compare it with the results obtained by inverse-Q filter methods. The maximum coefficient of a local crosscorrelation and its lag are used as attributes to estimate the similarity between the expected and the real output. Particularly the crosscorrelation lag is used as indicator of the phase correction. The uncertainty associated with the estimation of Q is included as variable by applying inverse-Q filter with a different Q from the one used to model the attenuated trace.
View full article as PDF (0.61 Mb)